Images AI Can't Generate?

I just thought about the human process of creating art. Images painted or drawn.

A human is perfectly able to stop half way. I wasn't able to create such an image with stable diffusion. The closest was a sketch pad with an half drawn image on it, but the image still was complete.

It seems to me that this is something an AI is not able to. And all the steps in between of how to create a drawing.

I know it is because it isn't trained to do that, but in my opinion it really show how little knowledge these systems actually have.

What do you think?